PROCESS DESCRIPTION In a limestone contactor, water flows through a bed of crushed sieved limestone in a similar way as it would flow through a sand filter (Spencer, 2000). The pH of water that flows through the limestone bed will be adjusted until it nears equilibrium with calcium carbonate (CaCO3(s)). The components of a contactor include a contact tank, limestone bed, inlet line, outlet line, overflow line, access lid (for closed system contactor only) and a flush outlet line (DeSouza et al., 2000). There are two types of contactors: (i) open and (ii) closed system contactor. The former is exposed to the atmosphere and the latter is covered from the atmosphere. There are also contactors that are built in pressurized vessels (Stauder, 2002). Limestone contactors are typically located at the end of the treatment train – after filtration, primary disinfection and chlorine contact (Spencer, 2000 and Benjamin et al., 1992). According to Spencer (2000) and Benjamin et al. (1992), by placing a contactor this way ensures the pH is maintained at a low level for effective chlorination. However, there are also contactors that are located after filtration and before chlorine addition (Stauder, 2002). Most of these contactors are located in Germany. The reason for placing the contactor before chlorine addition is because disinfection is not required in most limestone contactor plants in Germany especially if the plants are treating high quality deep spring water with very low turbidity (less than 0.1 NTU) and no or very low numberof coliforms (Stauder, 2003). The SSP takes a side stream of unstabilized water and doses it with carbon dioxide (De Souza et al., 2000). The acidified CO2-dosed side-stream then contacts a limestone bed, which will dissolve a considerable amount of CaCO3, increasing alkalinity and calcium concentration in the water. Much of the remaining CO2 is recovered by stripping with air and reused in the process whereas the stabilized side-stream blends with the main stream for full stabilization. The Simplified SSP is similar to the SSP but excludes CO2 recovery. In this process, there are two options that can be carried out after the acidified CO2-dosed side-stream contacts with a limestone bed: (1) Strip CO2 from the sidestream with no recovery, or (2) blend the sidestream and mainstream without stripping (De Souza et al., 2000). The Spraystab I is intended for small groundwater systems that need to remove iron or manganese. It combines aeration, limestone stabilization and filtration in one tank (Mackintosh, De Souza and De Villiers, 2003). The raw groundwater is first aerated to strip excess carbon dioxide from water and dissolve oxygen in the water to be stabilized. Finally, the water flows through a dual media filter (top layer of hydro-anthracite and lower layer of filter sand). In the filter, the limestone fines and other insoluble matter such as iron and manganese flocs are removed. The Spraystab II is intended for small groundwater systems that do not need to remove iron or manganese. It is similar to Spraystab I but excludes multi-media filtration and the water flows in the contactor in an upward direction (Mackintosh, De Souza and De Villiers, 2003). The raw groundwater is first aerated and then flows downward through a tube to the base of the limestone stabilization unit. Sludge and iron flocs (if iron is present in the water) will be collected at the base of the unit. Then, the water is collected and uniformly distributed by a slotted pipe manifold system through the limestone bed in an upward direction to be stabilized. (A) CHEMICAL REACTIONS INVOLVED IN LIMESTONE DISSOLUTION In order to understand the reactions involved in a limestone contactor, one must understand the basic principles governing the carbonate system in natural water and its equilibrium with limestone. Natural water contains carbonate species such as aqueous or - dissolved carbon dioxide (CO2(aq)), carbonic acid (H2CO3), bicarbonate (HCO3 ) and 2- carbonate (CO3 ) (Snoeyink and Jenkins, 1980). In a limestone contactor, the concentrations of the dissolved carbonate species are driven toward chemical equilibrium with CaCO3 by dissolving limestone. It is the interaction of these species that controls the pH in natural water (De Souza et. al., 2000) and can be undersaturated, in equilibrium or oversaturated with CaCO3 although low pH and alkalinity waters here are undersaturated with CaCO3. The degree of CaCO3 saturation of water is commonly calculated using the Langelier Index (L.I.). L.I. is the difference between the actual pH and hypothetical pH at equilibrium. Undersaturated water is represented by a negative value of L.I. and tends to dissolve CaCO3 whereas oversaturated water is represented by a positive value of L.I. and tends to precipitate CaCO3. Other procedures that can be used to determine the state of saturation of CaCO3 include evaluating the Calcium Carbonate Dissolution Potential, Ryznar (Saturation) Index, general saturation index also known as disequilibrium index and Larson's Ratio (Schock, 1999b). Marble Test and the carbonate saturometer device can also be used and both operate on the same principle (Schock, 1999b). If one assumes that the only aqueous species are H+, OH-, calcium and carbonate species, then reactions 1 to 7 take place when unstabilized water dissolves limestone in a closed system and reaches equilibrium. Equation 1. Dissociation of Water + − + − H 2O ⇔ H + OH KW = {H }{OH } Equation 2. Dissolution of CaCO3 from Limestone 2+ 2− 2+ 2− CaCO3(s) ⇔ Ca + CO3 Ksp = {Ca }{CO3 } - Equation 3. Formation of HCO3 − 2− + − {HCO3 } CO3 +H ⇔ HCO3 K a,2 = + 2− {}{}H CO3 * Equation 4. Formation of H2CO3 * − + * {H 2CO3 } HCO3 +H ⇔ H 2CO3 K a,1 = + − {}{}H HCO3 o Equation 5. Formation of CaCO3 Complex: o 2+ 2− o {CaCO3 } Ca +CO3 ⇔ CaCO3 K = 2+ 2− {}{}Ca CO3 Equation 6. Formation of CaOH+ Complex: CaOH + Ca 2+ +OH − ⇔ CaOH + K = {} {}{}Ca 2+ OH − + Equation 7. Formation of CaHCO3 Complex: + 2+ − + {CaHCO3 } Ca +HCO3 ⇔ CaHCO3 K = 2+ − {}{}Ca HCO3 In reactions 1 to 7, activities rather than concentrations are used in expressing the corresponding equilibrium constants. This is due to increased electrostatic interactions between ions as concentration of ions in solution increases (Snoeyink and Jenkins, 1980). As a result, the activity of ions becomes less than their measured or analytical concentration. Several mass balances must also be met as follows. Equation 8. Total Calcium Concentration 2+ + 0 + CT ,Ca = [Ca ]+ [CaHCO3 ]+ [CaCO3 ]+ [CaOH ] Equation 9. Total Carbonate Concentration * − + o 2− CT ,CO3 = [H 2CO3 ]+ [HCO3 ]+ [CaHCO3 ]+ [CaCO3 ]+ [CO3 ] Equation 10. Charge Balance 2+ + + + − 2− − 2[Ca ]+ [CaHCO3 ]+ [CaOH ]+ [H ]= [HCO3 ]+ 2[CO3 ]+ [OH ] Equation 11. Concentration Balance CT ,Ca = CT ,CO3 Calculations for real waters soon become more complicated because of the presence of other species but the principles are the same. Realistically, computer softwares such as AQUACHEM, MINEQL, MINTEQ, STASOFT or custom programs are used. Both Letterman and Kothari (1995) and Schott (2003) have developed programs specifically for solving dissolution of limestone in contactors as discussed later. (B) MATHEMATICAL MODEL OF LIMESTONE DISSOLUTION As CaCO3 dissolves from the limestone contactor media, particle size, bed depth, bed porosity, flow velocity and pressure drop change with time. It is important to make design choices for these variables since they affect the dissolution rate and recharge frequency. Letterman and Kothari (1995) and Haddad (1986) developed models of limestone dissolution rate that are sensitive to these variables and can be used for design. Haddad (1986) used Equation 12 and 13 to model the limestone dissolution process in a contactor operating at steady state. Equation 12. Calcium Carbonate Dissolution Rate, r r = ko a(Ceq − C) where: r = Calcium carbonate dissolution rate (moles/cm2.s). Ceq = Equilibrium calcium ion concentration, moles/L. C = Bulk calcium concentration, moles/L. -1 a = Area of CaCO3 per unit volume of fluid, cm . ko = Overall dissolution rate constant, cm/s. Equation 13. Continuity Equation d 2C dC N −θε + r = 0 D dZ 2 dZ where: N D = Axial dispersion number, dimensionless. Z = Depth, dimensionless. ε = Porosity of limestone particles, dimensionless. θ = Mean fluid residence time, sec. Using equations 12 and 13, Letterman, Haddad and Driscoll (1991) developed a steady- state model that relates the depth of limestone required in the contactor to the desired effluent water chemistry, influent water chemistry, limestone particle size and shape, bed porosity, water temperature and superficial velocity (Equation 14). Equation 14. Substitution of Equation 12 into 13 2 Ceq − CbL ko aLε ko aLε = exp− + N D C − C U U eq bo s s where: Cbo = Calcium concentration in the influent of a contactor, moles/L. CbL = Calcium concentration in the effluent of a contactor, moles/L. L = Overall depth of a contactor, cm. U s = Superficial velocity of fluid, cm/s. N D = Axial dispersion number, dimensionless. This model assumes that the rate of dissolution is controlled by two resistances that act in series: a surface reaction that controls the release of calcium from the solid and a mass transfer resistance that controls the rate of calcium transport between the solid surface and the bulk solution (Letterman and Kothari, 1995). Based on this kinetic model, Letterman and Kothari (1995) developed a computer program called DESCON. It can be downloaded from http://web.syr.edu/~rdletter/software.htm. It is used to facilitate the design of limestone contactors and will be discussed later in this module. (C) FACTORS AFFECTING CaCO3 DISSOLUTION RATE IN A LIMESTONE CONTACTOR Dissolution of limestone increases pH, alkalinity and dissolved inorganic carbon (DIC) of water and depletes the amount of limestone in a bed (Haddad, 1986). As the amount of limestone lessens over time, the bed depth and contact time will also be reduced.